A large proportion of methadone patients are dependent on multiple drugs. Alcohol is one drug that is frequently abused by methadone patients. The consequences of excessive alcohol use may include increased HIV risk-taking behaviors, psychosocial problems, accidental injury, and death. Despite the severity of the problem, a dearth of literature exists on treatment strategies for alcohol dependence in methadone patients. Only three controlled clinical trials have demonstrated significant reductions in alcohol use in this population. All three studies showed that an intervention in which disulfiram ingestion was necessary for methadone treatment reduced alcohol use compared to a standard, optional disulfiram condition. The two proposed studies will replicate and extend these findings. Specifically, Study 1 will assess whether disulfiram is necessary, or whether contingency management procedures on their own are sufficient, to reduce alcohol use, alcohol-related problems, and HIV risk-taking behaviors in alcohol dependent methadone patients. This study will compare three treatments: a treatment in which methadone dose is contingent upon disulfiram ingestion, a treatment in which methadone dose is contingent upon negative breath-alcohol samples (BACs), and a control treatment in which methadone dose is provided irrespective of BAC results. Study 2 will examine whether enhancing a treatment from Study 1 further improves outcomes. The study will compare: the most effective treatment in Study 1 (methadone contingent upon either disulfiram ingestion or negative BACs) and an enhanced version of that treatment in which both methadone dose and voucher incentives, exchangeable for retail goods, are contingent on either disulfiram ingestion or negative BACs. In the unlikely event that no differences are noted among the three conditions in Study 1, a treatment in which methadone dose and voucher incentives are contingent on disulfiram ingestion will be compared to the standard, non-contingent treatment. In total, 180 methadone maintained subjects will be randomly assigned to receive one treatment for a 3-month period. BACs will be assessed daily, urines will be screened for alcohol and illicit drugs twice weekly, and self-reports of HIV risk-taking behaviors will be collected weekly. Monthly interviews will assess psychiatric symptoms, employment and social functioning. Overall, these studies will (1) produce data on the relationship between alcohol consumption, other drug use, HIV risk- taking behaviors, and psychosocial functioning and (2) provide information that can be used by clinicians to treat alcohol dependence and alcohol-related problems among methadone patients.